Block copolymers comprise sequences (“blocks”) of the same monomer unit, covalently bound to sequences of unlike type. The blocks can be connected in a variety of ways, such as A-B in diblock and A-B-A triblock structures, where A represents one block and B represents a different block. In a multi-block copolymer, A and B can be connected in a number of different ways and be repeated multiply. It may further comprise additional blocks of different type. Multi-block copolymers can be either linear multi-block, multi-block star polymers (in which all blocks bond to the same atom or chemical moiety) or comb-like polymers where the B blocks are attached at one end to an A backbone.
A block copolymer is created when two or more polymer molecules of different chemical composition are covalently bonded to each other. While a wide variety of block copolymer architectures are possible, a number of block copolymers involve the covalent bonding of hard plastic blocks, which are substantially crystalline or glassy, to elastomeric blocks forming thermoplastic elastomers. Other block copolymers, such as rubber-rubber (elastomer-elastomer), glass-glass, and glass-crystalline block copolymers, are also possible.
One method to make block copolymers is to produce a “living polymer”. Unlike typical Ziegler-Natta polymerization processes, living polymerization processes involve only initiation and propagation steps and essentially lack chain terminating side reactions. This permits the synthesis of predetermined and well-controlled structures desired in a block copolymer. A polymer created in a “living” system can have a narrow or extremely narrow distribution of molecular weight and be essentially monodisperse (i.e., the molecular weight distribution is essentially one). Living catalyst systems are characterized by an initiation rate which is on the order of or exceeds the propagation rate, and the absence of termination or transfer reactions. In addition, these catalyst systems are characterized by the presence of a single type of active site. To produce a high yield of block copolymer in a polymerization process, such catalysts must exhibit living characteristics to a substantial extent.
Butadiene-isoprene block copolymers have been synthesized via anionic polymerization using the sequential monomer addition technique. In sequential addition, a certain amount of one of the monomers is contacted with the catalyst. Once a first such monomer has reacted to substantial extinction forming the first block, a certain amount of the second monomer or monomer species is introduced and allowed to react to form the second block. The process may be repeated using the same or other anionically polymerizable monomers. However, ethylene and other α-olefins, such as propylene, butene, 1-octene, etc., are not directly block polymerizable by anionic techniques.
Whenever crystallization occurs under quiescent conditions, which means that the polymer is not subjected to either external mechanical forces or unusually fast cooling, homopolymers made from a highly crystallizable monomer will crystallize from a melt and form spherical structures called “spherulites”. These spherulites range in size from micrometers to millimeters in diameter. A description of this phenomenon may be found in Woodward, A. E., Atlas of Polymer Morphology, Hanser Publishers, New York, 1988. The spherulites are composed of layer like crystallites called lamellae. Descriptions of this may be found in Keller, A., Sawada, S. Makromol. Chem., 74, 190 (1964) and Basset, D. C., Hodge, A. M., Olley, R. H., Proc. Roy. Soc. London, A377, p 25, 39, 61 (1981). The spherulitic structure starts from a core of parallel lamellae that subsequently branch and grow outward from the core in a radial direction. Disordered polymeric chains make up the material between lamellar branches as described in Li, L., Chan, C., Yeung, K. L., Li, J., Ng, K., Lei, Y., Macromolecules, 34, 316 (2001).
Polyethylene and random α-olefin copolymers of ethylene can be forced to assume non-spherulitic morphologies in certain cases. One situation occurs when the crystallization conditions are not quiescent, such as during blown or cast film processing. In both cases, the melts are subjected to strong external forces and fast cooling, which usually produce row-nucleated or “shish-kebab” structures as described in A. Keller, M. J. Machin, J. Macromol. Sci. Phys., 1, 41 (1967). A non-spherulitic morphology will also be obtained when the molecules contain enough of an α-olefin or another type of comonomer to prevent the formation of lamellae. This change in crystal type occurs because the comonomers are usually too bulky to pack within an ethylene crystal and, therefore, a sequence of ethylene units in between comonomers cannot form a crystal any thicker than the length of that sequence in an all-trans conformation. Eventually, the lamellae would have to become so thin that chain folding into lamellar structures is no longer favorable. In this case, fringed micellar or bundled crystals are observed as described in S. Bensason, J. Minick, A. Moet, S. Chum, A. Hiltner, E. Baer, J. Polym. Sci. B: Polym. Phys., 34, 1301 (1996). Studies of low molecular weight polyethylene fractions provide an understanding of the number of consecutive ethylene units that are required to form a chain folded lamellae. As described in L. Mandelkern, A. Prasad, R. G. Alamo, G. M. Stack, Macromolecules, 23, 3696 (1990) polymer chain segments of at least 100 ethylene units are required for chain folding. Below this number of ethylene units, low molecular weight fractions form extended chain crystals while polyethylene at typical molecular weights forms fringed micelles and creates a granular type morphology.
A fourth type of solid state polymer morphology has been observed in α-olefin block copolymers made by batch anionic polymerization of butadiene followed by hydrogenation of the resulting polymer. At the crystallization temperature of the ethylene segments, the amorphous blocks can be either glassy or elastic. Studies of crystallization within a glassy matrix have used styrene-ethylene (S-E) diblocks as described in Cohen, R. E., Cheng, P. L., Douzinas, K., Kofinas, P., Berney, C. V., Macromolecules, 23, 324 (1990) and ethylene-vinylcyclohexane (E-VCH) diblocks as described in Loo, Y. L., Register, R. A., Ryan, A. J., Dee G. T., Macromolecules 34, 8968 (2001). Crystallization within an elastic matrix has been studied using ethylene-(3-methyl-butene) diblocks as described in Quiram, D. J., Register, R. A., Marchand, G. R., Ryan, A. J., Macromolecules 30, 8338 (1997) and using ethylene-(styrene-ethylene-butene) diblocks as described in Loo, Y. L., Register, R. A., Ryan, A. J., Macromolecules 35, 2365 (2002). When the matrix was either glassy or was elastic but with a high degree of segregation between the blocks, the solid state structure showed the classical morphology of amorphous block copolymers such as styrene-butadiene-styrene (SBS), in which the different polymer segments were constrained into microdomains of approximately 25 nm in diameter. Crystallization of the ethylene segments in these systems was primarily constrained to the resulting microdomains. Microdomains can take the form of spheres, cylinders, lamellae, or other morphologies. The narrowest dimension of a microdomain, such as perpendicular to the plane of lamellae, is constrained to <60 nm in these systems. It is more typical to find constraints on the diameter of the spheres and cylinders, and the thickness of the lamellae to <30 nm Such materials may be referred to as microphase separated. FIG. 1 shows the predicted lamellar domain thickness for monodisperse ethylene/octene diblock copolymers at different values of total molecular weight and Δ octene mole %. The figure demonstrates that, even at very large differences in octene content of the blocks, molecular weights in excess of 180,000 g/mol are necessary to achieve domain sizes of 50 nm. The high viscosity which is unavoidable at such high molecular weights greatly complicates the production and processing of these materials. The calculation applied the theoretical results of Matsen, M. W.; Bates, F. S. Macromolecules (1996) 29, 1091 at a temperature of 140° C., a characteristic ratio of 7.5, and a melt density of 0.78 g/cm3. The correlation between octene mole % and χ was determined using the experimental results of Reichart, G. C. et al, Macromolecules (1998) 31, 7886.
Block copolymers containing both crystalline and amorphous blocks can crystallize from disordered, rather than microphase separated, melts and produce a regular arrangement of crystalline lamellae as described in Rangarajan, P., Register, R. A., Fetters, L. J. Macromolecules, 26, 4640 (1993). The lamellar thickness of these materials is controlled by the composition and molecular weight of both blocks as described in theories by Dimarzio, E. A., Guttmann, C. M., Hoffman, J. D., Macromolecules, 13, 1194 and Whitmore, M. D., Noolandi, J., Macromolecules, 21, 1482 (1988). For an ethylene based block copolymer, the maximum thickness of the crystalline region of these morphologies is the same as the maximum thickness of a high density polyethylene crystal which is about 22 nm.
These materials based on batch anionic polymerization can be additionally characterized as having very narrow molecular weight distributions, typically with Mw/Mn <1.2, and correspondingly narrow molecular weight distributions of their individual segments. They have also only been examined in the form of diblock and triblock copolymers since these are more readily synthesized via living anionic polymerization than structures with higher numbers of blocks.
Block copolymers from olefin monomers prepared using living polymerization catalysts were recently reviewed by Domski, G. J.; Rose, J. M.; Coates, G. W.; Bolig, A. D.; Brookhart, M., in Prog. Polym. Sci. 32, 30-92, (2007). Some of these monodisperse block copolymers also showed the classical morphology of amorphous block copolymers such as styrene-butadiene-styrene (SBS). Several of these block copolymers contain crystallizable segments or blocks, and crystallization of the segments in these systems was primarily constrained to the resulting microdomains. Syndiotactic polypropylene-block-poly(ethylene-co-propylene) and syndiotactic polypropylene-block-polyethylene, as described in Ruokolainen, J., Mezzenga, R., Fredrickson, G. H., Kramer, E. J., Hustad, P. D., and Coates, G. W., in Macromolecules, 38(3); 851-86023 (2005), form microphase separated morphologies with domain sizes consistent with monodisperse block copolymers (<60 nm). Similarly, polyethylene-block-poly(ethylene-co-propylene)s, as described by Matsugi, T.; Matsui, S.; Kojoh, S.; Takagi, Y.; Inoue, Y.; Nakano, T.; Fujita, T.; Kashiwa, N. in Macromolecules, 35(13); 4880-4887 (2002), are described as having microphase separated morphologies. Atactic polypropylene-block-poly(ethylene-co-propylene)s with narrow molecular weight distributions (Mw/Mn=1.07-1.3), as described in Fukui Y, Murata M. Appl. Catal. A 237, 1-10 (2002), are claimed to form microphase separated morphologies when blended with isotactic polypropylenes, with domains of amorphous poly(ethylene-co-propylene) between 50-100 nm. No microphase separation was observed in the bulk block copolymer.
Microphase separated diblock and triblock olefin block copolymers in which both block types are amorphous have also been prepared using living olefin polymerization techniques. A triblock poly(1-hexene)-block-poly(methylene-1,3-cyclopentene)-block-poly(1-hexene) copolymer, as described by Jayaratne K. C., Keaton R. J., Henningsen D. A., Sita L. R., J. Am. Chem. Soc. 122, 10490-10491 (2000), with Mn=30,900 g/mol and Mw/Mn=1.10 displayed a microphase separated morphology with cylinders of poly(methylene-1,3-cyclopentane) sized about 8 nm wide. Poly(methylene-1,3-cyclopentane-co-vinyltetramethylene)-block-poly(ethylene-co-norbornene) and poly(ethylene-co-propylene)-block-poly(ethylene-co-norbornene), as described by Yoon, J.; Mathers, R. T.; Coates, G. W.; Thomas, E. L. in Macromolecules, 39(5), 1913-1919 (2006), also display microphase separated morphologies. The poly(methylene-1,3-cyclopentane-co-vinyltetramethylene)-block-poly(ethylene-co-norbornene), with Mn=450,000 g/mol and Mw/Mn=1.41, has alternating domains of 68 and 102 nm, while the poly(ethylene-co-propylene)-block-poly(ethylene-co-norbornene), with Mn=576,000 g/mol and Mw/Mn=1.13, has domains sized 35-56 nm. These samples demonstrate the difficulty in achieving domain sizes >60 nm, as very high molecular weights are required to achieve such large domains.
Achieving microphase separated block copolymer morphologies usually requires unfavorable dispersive interactions between the segments of the different blocks, as characterized by the Flory-Huggins χ parameter, and high molecular weights. Representing the average block molecular weight as N, a typical narrow polydispersity diblock containing equal amounts by volume of the two blocks requires a value of χ times N greater than 5.25 for the melt to display an ordered microphase morphology as shown by L. Leibler, Macromolecules 13, 1602 (1980). This minimum value of χN to achieve order increases to about 6 for triblock copolymers with equal volumes of the two block types. As the number of blocks per molecule increases further, the required χN also increases and asymptotically approaches 7.55 in the limit of a large number of blocks per molecule as shown by T. A. Kavassalis, M. D. Whitmore, Macromolecules 24, 5340 (1991). Although multiblocks such as pentablocks have been shown to provide a substantial improvement in mechanical properties as described in T. J. Hermel, S. F. Hahn, K. A. Chaffin, W. W. Gerberich, F. S. Bates, Macromolecules 36, 2190 (2003), the overall molecular weight of these multiblocks has to be large in order to meet the requirements for ordered melt morphologies. Since the energy requirements to process a polymer increase sharply with molecular weight, the commercial opportunities of such multiblocks may be limited.
However, theoretical studies by S. W. Sides, G. H. Fredrickson, J. Chem. Phys. 121, 4974 (2004) and D. M. Cooke, A. Shi, Macromolecules 39, 6661 (2006) have shown that the minimum χN for ordered morphologies decreases as the polydispersity of one or both of the block types is increased. When both block types have a most probable distribution of length, i.e. the ratio of weight average to number average block molecular weight is 2, the minimum value of χN, where N is the number average block length, in order to achieve an ordered morphology is 2 for equal volumes of the two block types as shown by I. I. Potemkin, S. V. Panyukov, Phys. Rev. E. 57, 6902 (1998) for multiblocks in the mean-field limit. This lower value in χN translates to a substantial reduction in overall molecular weight for a melt ordered multiblock and, therefore, a drop in processing costs.
Another prediction of interest made by Potemkin, Panyukov and also by Matsen, M. W., Phys. Rev. Lett. 99, 148304 (2007) is that each transition in morphology, including the transition from disorder to order, does not occur abruptly as in monodisperse block copolymers. Instead, there are regions of coexisting phases along each boundary. Along the order-order boundaries, the overall composition of a molecule may determine how it partitions between phases. For example, polydisperse diblocks along the boundary between cylindrical and lamellar phases may have the more symmetric diblocks form lamellae while the asymmetric ones will tend to form cylinders. In the vicinity of the order-disorder boundary, molecules with longer blocks may form an ordered morphology while those with shorter blocks remain disordered. In some cases, these disordered molecules may form a distinct macrophase. Alternatively, the location of these molecules could be directed to the center of the ordered domains in a similar manner to the domain swelling that occurs when a homopolymer is blended with a block copolymer (Matsen, M. W., Macromolecules 28, 5765 (1995)).
In addition to achieving microphase separation at lower values of χN, block length polydispersity has also been hypothesized to have a pronounced effect on the domain spacing of the ordered structures. The size of the microdomains in monodisperse block copolymers is largely a function of the average molecular weight of a block, N, and is typically on the order of ˜20-50 nm. However, it has been predicted that polydispersity leads to larger domain spacings as compared with equivalent monodisperse block copolymers (Cooke, D. M.; Shi, A. C. Macromolecules, 39, 6661-6671 (2006); Matsen, M. W., Eur. Phys. J. E, 21, 199-207 (2006)). The effects of polydispersity on phase behavior have also been demonstrated experimentally. Matsushita and coworkers approximated polydispersity by blending a series of monodisperse polystyrene-b-poly(2-vinylpyridine)s (Noro, A.; Cho, D.; Takano, A.; Matsushita, Y. Macromolecules, 38, 4371-4376 (2005)). Register and coworkers found ordered morphologies in a series of polystyrene-b-poly(acrylic acid)s synthesized using a controlled radical polymerization technique (Bendejacq, D.; Ponsinet, V.; Joanicot, M.; Loo, Y. L.; Register, R. A. Macromolecules, 35, 6645-6649 (2002)). Most recently, Lynd and Hillmyer (Lynd, N. A.; Hillmyer, M. A. Macromolecules, 38, 8803-8810 (2005)) evaluated a series of monodisperse poly(ethylene-alt-propylene)s that were chain extended with a block of poly(DL-lactide) using synthetic techniques that introduced polydispersity in the poly(DL-lactide) block. In all of these examples polydispersity led to increased domain spacings, suggesting that the longer blocks have a greater role in determining domain size. In some instances, polydispersity also produced changes in the type of ordered morphology. The range of techniques for synthesis of polydisperse block copolymers is extremely limited, and it is especially difficult to introduce polydispersity in multiple blocks while maintaining a high fraction of block copolymer.
The tendency for longer block lengths to have a greater role in determining domain size, combined with the ability to swell domains, creates the potential for domain sizes that are much larger than what is observed in typical monodisperse block copolymers. The ability for some molecules to be ordered and others disordered contributes to the formation of swollen domains. The morphology of these systems can be termed block copolymer-directed mesophase separation.
It would be useful to provide an olefin block copolymer with an overall molecular weight distribution and segment molecular weight distribution such that Mw/Mn >1.4, that is mesophase separated. It would also be useful to provide such a material that is a multi-block copolymer with a distribution in the number of blocks.
In addition, there is an unfulfilled need for mesophase separated block copolymers which are based on ethylene and α-olefins. There is also a need for block copolymers with low molecular weights (Mw<250,000 g/mol) that form domains larger than those from monodisperse block copolymers of the prior art, namely greater than 60 nm in the smallest dimension. There is also a need for a method of making such block copolymers.